When AI decides how shareholders vote, boards need to rethink governance

A major asset manager has replaced external proxy advisory firms with an internal AI system to drive proxy voting, shifting fiduciary judgment upstream into model design, training data and override protocols. The change raises governance risks for corporate boards—AI can misread nuance, aggregate signals continuously, and produce votes without clear appeals—so boards should strengthen narrative disclosures and investor engagement to ensure context and correctability before votes are cast.

Analysis

Market structure: Expect winners that sell governance workflow, disclosure structuring, and internal stewardship systems (Broadridge BR, Workiva WK, and large custodians) as asset managers internalize voting. Pure-play external proxy advisers (private ISS/Glass Lewis) risk 20–40% revenue erosion over 2–3 years as top 10 asset managers divert volume; pricing power will shift to integrated vendors who can supply auditable AI pipelines.

Risk assessment: Key tail risks include regulatory intervention (SEC/EU requiring transparency or audit trails for AI voting within 6–18 months), systemic misvotes triggering class actions, and model-data poisoning events; any of these could cause >15% moves in affected equities and widen HY credit spreads by 50–150bps in governance-fragile issuers. Hidden dependencies: voting algorithms rely on noisy public text — disclosure ambiguity will create persistent event-driven volatility around proxy windows.

Trade implications: Favor 6–12 month overweight to BR (2–3% portfolio) and WK (1–2%) to capture secular demand for structured governance disclosures and proxy infrastructure; hedge with a 3-month S&P put spread (5%/8% strikes) sized to cost <0.5% portfolio to protect event risk. Add modest core long to BLK (1–2%) for scale advantage in building internal stewardship; consider 9–12 month call spreads instead of outright if capital efficiency desired.

Contrarian angle: The market underestimates incumbent adaptation — proxy advisors will sell AI tooling/licensing rather than disappear, creating arbitrage between SaaS vendors (BR/WK) and advisory-only models. Also, increased algorithmic screening could paradoxically raise activist activity by making targets easier to find, producing trading opportunities in credit (buy protection on single-name bonds with governance red flags) not yet priced in.

Key Decisions for Investors

• Allocate 2–3% portfolio long to Broadridge Financial Solutions (BR) over 6–12 months to capture incremental proxy processing and governance-data services; target +20–30% upside, stop-loss 12% below entry.
• Establish 1–2% long position in Workiva (WK) over 3–9 months to benefit from demand for structured governance disclosures; target +15–25% upside, stop-loss 15%.
• Buy a 3-month S&P500 put spread (5%/8% strikes) sized to cost <=0.5% of portfolio as protection against event-driven proxy-season volatility and potential governance shocks.
• Add 1–2% long to BlackRock (BLK) or a similarly scaled asset manager as a convex play on internal stewardship scale; if SEC/EU publishes AI-voting transparency rules within 90 days, rotate 0.5–1% from BLK into BR to favor infrastructure over asset-gathering exposure.